1 code implementation • ICCV 2023 • Dongyue Chen, Tingxuan Huang, Zhimin Song, Shizhuo Deng, Tong Jia
In the encoding stage, an Attention Guided Gated-Convolution (AG-GConv) module is proposed to realize the fusion of depth and color features at different scales, which can effectively reduce the negative impacts of invalid depth data on the reconstruction.